SShortSingh.
TopTechnologyAI & MLStartupsProgrammingScienceBusinessFinanceCrypto & Web3World

Programming

LatestTopDiscussed
0
ProgrammingDEV Community ·

How Munchable Built a Smarter Postgres Backup Job That Fails With Clear Errors

Munchable, whose Postgres database lacks managed backups, relies entirely on a nightly GitHub Actions job to dump data to Cloudflare R2 and alert admins on failure. The backup is split into two archives: a full dump retained for 30 days and a smaller core dump — excluding the rebuildable product catalogue — kept for a year at near-zero cost within R2's free tier. After the first run failed with an uninformative 'Network is unreachable' error, developers added preflight checks that detect and explain common misconfigurations, such as using an IPv6-only direct host or the transaction pooler port that pg_dump cannot use. Each error message names the specific fix and points to a runbook file, so on-call engineers get actionable guidance rather than a bare failure. Bucket lifecycle rules, not the workflow itself, handle backup deletion to avoid accidental data loss from scripting bugs.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Renaming a launchd Plist File Does Not Stop the Job — Here Is Why

A developer managing over 160 macOS launchd automation jobs discovered that renaming a plist file to a '.retired' extension does not actually stop the job from running. This happens because launchd identifies jobs by the Label key inside the plist file, not by the file's name or path, so once a job is loaded into launchd's internal table it continues to run regardless of any rename. Unlike Linux's systemd, which ties service identity to the unit file path, macOS launchd treats the file path only as an initial loading source and manages everything thereafter via the Label value. A rename is interpreted by the OS as a delete-and-add event, but files without a .plist extension are ignored by launchd's directory watcher on macOS 13 and later, meaning the already-loaded job entry is never automatically removed. On machines that run for long periods without a reboot, this behavior can cause supposedly retired scheduled jobs to keep executing silently in the background.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

How One Team Fixed Broken Event Sourcing in AWS With DynamoDB and SNS

A development team discovered critical flaws in their distributed event sourcing architecture, where application state and events were written separately rather than within a single transaction, risking data inconsistency. Their original microservices also published coarse-grained events containing full state snapshots, making it difficult to determine what actually changed and why. To fix this, the team introduced three dedicated DynamoDB tables for domain events, current application state, and integration events, using DynamoDB Streams to trigger SNS notifications with the necessary metadata. They replaced broad change events with granular, business-oriented events such as AccountCreated and EmailAddressReplaced, each carrying only the changed delta and a revision number. Rather than relying on FIFO queues to enforce message ordering, the team opted to expose revision numbers and let consumers handle resequencing themselves, acknowledging that strict ordering guarantees break down across complex distributed systems.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Developer rebuilds entire automation workflow after rogue bot switched git branch mid-deploy

A developer discovered a critical flaw in their automation setup when a scheduled bot silently switched git branches while a deploy was still running, causing the wrong build to land on the device. The root cause was multiple automated workflows sharing the same working directory, creating collisions when two processes ran simultaneously or overlapped with manual work. The fix involved adopting git worktree, a feature available since git 2.5 that lets each automation operate in its own isolated folder while sharing a single .git object store. The developer rebuilt all 10 automation workflows around this approach, ensuring no bot can ever interfere with the directory a human is actively using. Each workflow now follows a five-step pattern: trigger, isolate, work, verify, and notify without committing, keeping human judgment as the final step.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Specialized AI agent Mano-CUA outperforms Claude and Gemini on browser automation benchmark

A specialized vision-based AI agent called Mano-CUA 1.1 scored 41.7 on WebRetriever Protocol I, a benchmark testing real-world browser navigation and form-filling tasks. By comparison, Gemini 2.5 Pro Computer Use scored 40.9 and Claude 4.5 Computer Use scored 31.3, highlighting a notable performance gap. Unlike general-purpose models that rely on DOM parsing or accessibility APIs, Mano-P uses a pure vision-driven approach, interpreting screenshots the way a human would rather than reading page structure. This method proves more resilient on complex, dynamic web applications where page layouts shift after each user interaction. The model can run entirely on-device on Apple Silicon hardware with 32GB RAM, keeping user data local without sending screenshots to external servers.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Couple Builds AI Trip Planner Escapo to Simplify Travel Itinerary Planning

A couple frustrated with the time-consuming nature of travel planning built Escapo, a web-based trip planning tool available at escapo.io. Users begin by answering a short questionnaire about their destination, travel dates, and preferences, after which the app generates a personalized day-by-day itinerary using real, named locations. The itinerary is geographically optimized, grouping nearby places together with smart timing to reflect a realistic daily plan. An interactive map lets users visualize the full trip, and the final itinerary can be exported as a PDF or as map pins for offline use. The tool is currently free to try and was designed with the goal of reducing trip planning from hours to just a few minutes.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Modal Labs Raises $355M Series C, Valued at $4.65B Amid AI Infrastructure Boom

Modal Labs, a San Francisco-based cloud infrastructure startup founded by Erik Bernhardsson, closed a $355 million Series C funding round led by General Catalyst and Redpoint Ventures in May 2026. The deal values the company at $4.65 billion, roughly four times its previous valuation. Modal offers a serverless, Python-optimized compute platform that lets developers deploy AI workloads at scale without managing Kubernetes clusters or Docker containers. The company targets data scientists, ML engineers, and AI researchers who need rapid access to GPU resources for tasks ranging from model fine-tuning to large parallel training runs. The funding surge reflects rising demand for scalable AI infrastructure, driven in part by the widespread adoption of AI coding tools that generate compute-intensive applications.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Munchable builds 450-word test suite to prevent silent gaps in food label scanning

Food app Munchable discovered its ingredient-scanning rules engine was returning 'not assessed' for common condition-relevant ingredients, including kefir for a lactose-intolerant user, soya for an IBS user, and tea for a reflux user. Each failure traced back to a missing alias rather than an absent rule, meaning the engine could not match a label's spelling to the rule it was supposed to trigger. To prevent recurrence, the team created a hand-maintained file of 450 label words grouped by dietary condition, covering FODMAP, lactose, GERD, IBD, and gastroparesis categories. Each word in the list is tested by walking the same canonicalisation and taxonomy path that a real product scan follows, ensuring the promise in the file reflects actual engine behaviour. A separate list of 15 everyday words such as water, salt, and sugar is also checked, to stop the app from flagging harmless ingredients as unassessed and drowning out genuinely important alerts.

0 comments Read more at DEV Community
0
ProgrammingHacker News ·

Neovim holds ~$800k in untouched Bitcoin donations since at least 2023

A user discovered that the Neovim open-source text editor project has approximately 10 Bitcoin sitting in its public donation wallet, currently valued at around $800,000. The donation appears to have been made in 2023 and has remained untouched since then. Blockchain records suggest the project last moved funds out of the address in 2019, meaning the wallet may have gone unmonitored for several years. The discovery was shared on Hacker News, where the poster questioned whether the Neovim team was aware of the significant sum. The community expressed hope that the funds could be put to productive use for the project's development.

0 comments Read more at Hacker News
0
ProgrammingDEV Community ·

How Munchable built an allergen system that can only warn, never falsely clear

Munchable, an app helping people with IBS or reflux assess supermarket products, expanded in September 2024 to also flag declared allergens. Unlike its dietary condition logic, the allergen layer is designed to only add warnings and never mark a product as allergen-free, since absence of data is treated as silence rather than safety. This rule is enforced at three levels — in code logic, in result structure, and in a TypeScript type that contains no value for 'absent'. The system covers all 14 EU and UK regulated allergens, a superset of the US 'big nine', so warnings are not lost when users travel across regions. Careful distinctions are also made to avoid false positives, such as not treating an inferred generic 'nut' tag as evidence of peanuts unless that word appeared on the physical label.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

AI Debugging Tools Can Speed Up Frontend Fixes but Carry Real Risks

AI-powered frontend debugging tools work by parsing source code and collecting runtime data, then feeding that context into large language models like GPT-4 to suggest bug fixes or debugging steps. In a tested case, one such tool correctly identified a direct state mutation in a React component that was preventing re-renders, saving multiple rounds of manual trial-and-error. However, these tools have notable limitations, including a tendency to hallucinate confident but incorrect suggestions, such as flagging a non-existent missing import. They also lack true debugger access, meaning they can miss runtime side effects and may address symptoms rather than root causes. Developers are advised to treat AI suggestions as hypotheses, always review generated fixes, and combine AI insights with traditional tools like breakpoints and network inspectors.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

EU Cyber Resilience Act Reporting Rules Now Live: What Developers Must Know

The EU Cyber Resilience Act's (CRA) mandatory vulnerability reporting obligations came into effect on 11 September 2026, applying to any developer or company that commercially supplies software or hardware to the EU market. Under Article 14, manufacturers must notify ENISA and the relevant national CSIRT within 24 hours of discovering an actively exploited vulnerability, follow up with a detailed notification within 72 hours, and submit a final report within 14 days of a fix being available. The rules apply equally to individual developers selling apps or games as to large corporations, regardless of where the seller is based. Commercially active is defined broadly to include paid products, ad-supported software, and data-monetised services, not just direct sales. Broader CRA requirements — including SBOM documentation, CE marking, and a minimum five-year security support period — do not take effect until 11 December 2027.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

CompTIA Network+ Guide: Key Routing, Switching, and VLAN Concepts Explained

A detailed technical overview covers core Layer 2 and Layer 3 networking concepts relevant to the CompTIA Network+ (N10-009) certification exam. The guide explains static and dynamic routing methods, detailing how routers select the best path using longest prefix match, administrative distance, and metrics. It also covers Network Address Translation (NAT) types, including PAT, which maps multiple private IPs to a single public IP using port tracking. First Hop Redundancy Protocols such as HSRP, VRRP, and GLBP are outlined as solutions to eliminate single points of failure at the default gateway. Additionally, the article introduces VLANs as a way to logically segment a physical switch into isolated broadcast domains without additional hardware.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Developer launches free URL2Vid tool to download videos from 10+ platforms

A developer named Purushottam Barai has built and released URL2Vid, a free web app that allows users to download videos or audio from over ten platforms including YouTube, Instagram, TikTok, and Twitter/X by simply pasting a URL. The tool supports MP4, muted MP4, and MP3 output formats, and requires no account creation, app installation, or watermarks. Built with React, Node.js, and the yt-dlp extraction library, the project was motivated by the developer's frustration with ad-heavy and unreliable existing downloader sites. Key technical challenges included handling multiple URL formats across different platforms, such as YouTube Shorts and Reddit share-links, which were addressed through a comprehensive automated test suite. The project is in its early stages, with batch downloads and improved discoverability planned as future enhancements.

0 comments Read more at DEV Community
0
ProgrammingHacker News ·

Manticore Search Introduces Auto-Chunking to Improve Vector Search on Long Documents

Manticore Search has published a blog post detailing a new auto-chunking feature designed to enhance vector search performance on long documents. The feature addresses a common limitation in vector search systems, where lengthy texts are difficult to process accurately as a single unit. By automatically splitting documents into smaller chunks internally, Manticore Search aims to improve retrieval relevance without requiring manual preprocessing by developers. The announcement was shared on Hacker News, drawing modest early attention with five points and no comments at the time of posting.

0 comments Read more at Hacker News
0
ProgrammingDEV Community ·

Four Core Approaches That Make Deep Generative Models Work

A technical explainer published on DEV Community breaks down how deep generative models solve the challenge of representing complex, high-dimensional data distributions such as images and audio. Rather than treating generation as a single problem, different model families restructure it in distinct ways: Autoregressive Models decompose joint distributions into conditionals, Variational Autoencoders use latent variables, and Flow-based Models apply invertible transformations to simple distributions. Generative Adversarial Networks take a likelihood-free route, learning to generate data without directly computing probability densities. Most other families — including Diffusion and Score-based Models — fall under the likelihood-based category, where training is framed as minimizing Negative Log-Likelihood, derived from KL Divergence. The article emphasizes that the core challenge in generative modeling is not sample generation itself, but turning an intractable distribution into something a model can compute and optimize.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

How to generate AWS S3 presigned URLs in .NET without third-party SDKs

Developers building .NET applications often rely on the AWSSDK.S3 NuGet package to generate presigned S3 upload URLs, but this pulls in a large SDK that may be unnecessary for simple use cases. A lightweight alternative involves implementing the AWS Signature Version 4 signing process directly in code using built-in .NET cryptography libraries. The approach works with AWS S3 and compatible storage providers such as Cloudflare R2, Backblaze B2, DigitalOcean Spaces, and Hetzner Object Storage. The method generates a time-limited PUT URL with locked-in content type and file size constraints, requiring only the access key credentials to be stored as application secrets. This technique reduces external dependencies while still enabling secure, temporary file upload links for end users.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

LeaDQ Uses Reinforcement Learning to Select Unlabeled Samples for Federated Training

A research paper introduces LeaDQ, a method designed to solve the data-querying problem in federated learning, where clients continuously receive unlabeled data streams. Because labeling every incoming sample is costly, clients must decide which samples are worth spending their limited annotation budget on. The core challenge is that samples most useful to a local client may not be the most valuable for optimizing the shared global model. LeaDQ frames this selection process as a reinforcement learning problem, where a policy makes binary decisions on each incoming sample using the model's predictive logits as observations. Selected samples are sent to an expert for labeling, incorporated into federated training, and the cycle repeats as new unlabeled data arrives.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

KYB vs KYC: How Businesses Are Verified Beyond Individual Identity Checks

Know Your Business (KYB) and Know Your Customer (KYC) are often grouped together in compliance systems, but they address fundamentally different verification challenges. While KYC validates an individual's identity at a single point in time, KYB requires querying government registries, traversing corporate ownership structures, and identifying the real human beings who ultimately own or control a legal entity. A compliant KYB pipeline must screen both corporate and individual nodes against sanctions lists, politically exposed persons databases, and adverse media watchlists. If any linked individual fails an anti-money laundering or sanctions check, the entire entity-level KYB status is flagged as blocked. Without automated KYB processes, bad actors barred under individual KYC rules can potentially hide behind corporate structures to evade detection.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

How to Build a Reliable PCBA Inspection Feedback Loop That Drives Real Fixes

Effective printed circuit board assembly (PCBA) inspection requires linking data from stencil, paste, placement, and reflow stages into a single traceable record per panel. Defect records must distinguish observations from root-cause hypotheses to prevent unverified guesses from becoming accepted facts in dashboards. Teams are advised to pilot the process on one stable product with a limited feature set, freezing inspection parameters before collecting baseline data across varied shifts, material lots, and equipment states. Corrective actions should only be closed after a controlled production run confirms the fix without shifting failures elsewhere. A compact, structured data format and regular owner reviews with named experiments and success metrics are recommended to make local improvements transferable and auditable.

0 comments Read more at DEV Community
← NewerPage 1178 of 3224Older →